Fault diagnosis of spur gearbox based on random forest and wavelet packet decomposition

Abstract

This paper addresses the development of a random forest classifier for the multi-class fault diagnosis in spur gearboxes. The vibration signal’s condition parameters are first extracted by applying the wavelet packet decomposition with multiple mother wavelets, and the coefficients’ energy content for terminal nodes is used as the input feature for the classification problem. Then, a study through the parameters’ space to find the best values for the number of trees and the number of random features is performed. In this way, the best set of mother wavelets for the application is identified and the best features are selected through the internal ranking of the random forest classifier. The results show that the proposed method reached 98.68% in classification accuracy, and high efficiency and robustness in the models.

Keywords

fault diagnosis / spur gearbox / wavelet packet decomposition / random forest

References

[1] Walha L, Fakhfakh T, Haddar M. Backlash effect on dynamic analysis of a two-stage spur gear system. Journal of Failure Analysis and Prevention, 2006, 6(3): 60–68

[2] Abbes M S, Fakhfakh T, Haddar M,  Effect of transmission error on the dynamic behaviour of gearbox housing. International Journal of Advanced Manufacturing Technology, 2007, 34(3–4): 211–218

[3] Tian Z, Zuo M, Wu S. Crack propagation assessment for spur gears using model-based analysis and simulation. Journal of Intelligent Manufacturing, 2012, 23(2): 239–253

[4] Ebersbach S, Peng Z. Fault diagnosis of gearbox based on monitoring of lubricants, wear debris, and vibration. In: Wang Q, Chung Y W, eds. Encyclopedia of Tribology. New York: Springer, 2013, 1059–1064 

[5] Rgeai M, Gu F, Ball A,  Gearbox fault detection using spectrum analysis of the drive motor current signal. In: Kiritsis D, Emmanouilidis C, Koronios A, , eds. Engineering Asset Lifecycle Management. London: Springer, 2010, 758–769

[6] Hong L, Dhupia J S. A time domain approach to diagnose gearbox fault based on measured vibration signals. Journal of Sound and Vibration, 2014, 333(7): 2164–2180

[7] Rafiee J, Arvani F, Harifi A,  Intelligent condition monitoring of a gearbox using artificial neural network. Mechanical Systems and Signal Processing, 2007, 21(4): 1746–1754

[8] Sanchez R, Arpi A, Minchala L. Fault identification and classification of spur gearbox with feed forward back propagation artificial neural network. In: Proceedings of the 2012 Andean Region International Conference. Washington, D.C.: IEEE, 2012, 215 

[9] Barakat M, Lefebvre D, Khalil M,  Parameter selection algorithm with self-adaptive growing neural network classifier for diagnosis issues. International Journal of Machine Learning and Cybernetics, 2013, 4(3): 217–233

[10] Yang B S, Han T, An J L. ART-KOHONEN neural network for fault diagnosis of rotating machinery. Mechanical Systems and Signal Processing, 2004, 18(3): 645–657

[11] Jiang Z, Fu H, Li L. Support vector machine for mechanical faults classification. Journal of Zhejiang University SCIENCE A, 2005, 6(5): 433–439

[12] Jiao B, Xu Z. Multi-classification LSSVM application in fault diagnosis of wind power gearbox. In: Zhang T, ed. Mechanical Engineering  and  Technology.  Berlin:  Springer,  2012,  125:  277–283

[13] Kang Y, Wang C, Chang Y. Gear fault diagnosis in time domains by using Bayesian networks. In: Melin P, Castillo O, Ramirez E, , eds. Analysis and Design of Intelligent Systems using Soft Computing Techniques. Berlin: Springer, 2007, 41: 618–627

[14] Breiman L, Friedman J, Olshen R,  Classification and regression trees. The Wadsworth and Brooks-Cole statistics-probability series. Boca Raton: Chapman & Hall, 1984

[15] Breiman L. Random forests. Machine Learning, 2001, 45(1): 5–32

[16] Criminisi A, Shotton J. Classification forests. In: Criminisi A, Shotton J, eds. Decision Forests for Computer Vision and Medical Image Analysis. London: Springer, 2013, 25–45

[17] Han X, Yang B S, Lee S J. Application of random forest algorithm in machine fault diagnosis. In: Mathew J, Kennedy J, Ma L, , eds. Engineering Asset Management. London: Springer, 2006, 779–784 

[18] Yang B S, Di X, Han T. Random forests classifier for machine fault diagnosis. Journal of Mechanical Science and Technology, 2008, 22(9): 1716–1725

[19] Karabadji N, Khelf I, Seridi H,  Genetic optimization of decision tree choice for fault diagnosis in an industrial ventilator. In: Fakhfakh T, Bartelmus W, Chaari F, , eds. Condition Monitoring of Machinery in Non-Stationary Operations. Berlin: Springer, 2012, 277–283 

Acknowledgments

The authors want to express a deep gratitude to the Ministry of Higher Education, Science, Technology and Innovation of the Republic of Ecuador (SENESCYT), for their support with this research work. Fernando Sancho wants to thank to the Prometeo Project from the same institution, and the Project TIC-6064 from Junta de Andalucía (Spain) for their support.

RIGHTS & PERMISSIONS

2014 Higher Education Press and Springer-Verlag Berlin Heidelberg